|May 30, 2019|
|1 year 7 months ||Taking care of the case where the chosen time step is smaller than,
or equal to, 60 seconds, in which case adaptive time stepping should
not be active.
|1 year 7 months ||Making it default to redo the heat equation with a smaller time step
upon non-convergence, and use the key ALLOW_ADAPTIVE_TIMESTEPPING
exclusively to prematurely reduce the time step to reduce spurious
oscillations near the surface. So setting ALLOW…
|May 28, 2019|
|1 year 7 months ||For sea ice module: tracking total flooding when using the Bucket
|Apr 16, 2019|
|1 year 9 months ||LEHNING_1 albedo could run into a floating point exception in very
wet conditions, due to Alb1 < -1.
|1 year 9 months ||Typo.
|Apr 15, 2019|
|1 year 9 months ||Improving consistency in dealing with rounding errors in phase
|Apr 12, 2019|
|1 year 9 months ||Adding the missing case of mk 9 for a water layer. Note that
otherwise, a<0, such that the cast to unsigned short int
experiences an underflow and type >> 1000.
|1 year 9 months ||Improving documentation.
|Apr 9, 2019|
|1 year 9 months ||Setting initial snow salinity to 0 and disable the zero flux upper
boundary condition for advection. Both modifications are for
numerical stability, but it is not the desired setting for the model.
|1 year 9 months ||Adding comments to understand the function call.
|1 year 9 months ||Introducing stricter limits on the van Genuchten parameters for sea
ice, for the sake of numerical stability. Additionally removing some
|1 year 9 months ||Sea ice: introducing special condition to favour only evaporation,
not necessarily condensation when solving richards equation, to
|1 year 9 months ||For sea ice: when mass is removed from the domain, adjust the
pressure head in saturated conditions, to reflect the new pressure
|Apr 3, 2019|
|1 year 9 months ||Correct saturated pressure heads in case of element splitting.
|Apr 1, 2019|
|1 year 9 months ||Improvements with global (i.e., class-wide) switches for the
boundary conditions, and attempts to improve the explicit scheme,
but the results of that remain somewhat unsatisfying.
|1 year 9 months ||Measured snow height could get below ground level when it was in
fact not available. But it's not clear why measured snow height
needs to be corrected, as in the next time step, the new measured hs
value will be assigned. So only routines following …
|Mar 29, 2019|
|1 year 9 months ||Better search strategies
|1 year 9 months ||Better search priorities
|Mar 21, 2019|
|1 year 10 months ||Decreasing convergence thresholds for Richards solver, to improve
the stability of sea ice simulations.
|1 year 10 months ||Bug fix for SafeMode (SafeMode 1 and 2 could never work because
mass1 was set to 0 and never recalculated). Also slight improvement.